Abstract
In Barcelona, Spain prior to 2006, HIV testing was mostly limited to formal healthcare facilities with no incidence data reported. A community-based organization (BCN Checkpoint) was established to increase HIV testing in a peer-led community location to generate incidence data in men who have sex with men and transgender women. Three community engagement interventions were conducted between 2009 and 2017 as follows: 2009–2011 (peer-led point-of-care testing for HIV), 2012–2014 (12-monthly HIV testing with an emphasis on testing in partnerships), 2015–2017 (three-monthly HIV testing with rapid referral for antiretroviral initiation). Between 2009 and 2017 a predominantly cisgender male (99.4%) and Spanish national (62.4%) population with mean age of 34.8 years had 49,630 visits. Mean visit number increased from 1.69 in the first to 2.07 in the last three-year period. HIV incidence fell from 4.17 (95% confidence interval [CI]: 3.53–4.93) per 100 person-years in 2009–2011 to 1.57 (95% CI: 1.30–1.89) per 100 person-years in 2015–2017. This represents a 62% reduction (incidence rate ratio: 0.38, 95% CI: 0.29–0.48) between the first and third study period (p < 0.001). These early interventions may have contributed to the reduction seen in HIV incidence in this cohort.
Introduction
The city of Barcelona has a highly visible gay community that has been severely affected by the HIV/AIDS epidemic since the early 1980s. Despite this, there have been no formal statistics published on HIV incidence focused on men who have sex with men (MSM)/transgender women (TW) over the course of the epidemic.
In the years prior to BCN Checkpoint’s opening, the number of new cases of HIV reported in Catalonia ranged between 700 and 800 persons (rate: 11.7 cases per 100,000 inhabitants), almost all in male Spanish nationals: 85% were MSM, and 70% were between the ages of 20 and 40 years. Over 80% of cases were from the city of Barcelona with approximately 50% of new HIV infections from the “Eixample” area. AIDS notifications closely followed HIV case demographics and locations. 1
While the above data are helpful at a macro level to focus attention on the demographic at highest risk of acquiring HIV infection and where they are located, it does not provide adequate information by which the efficacy of interventions to reduce HIV incidence may be assessed.
As an easily accessible venue for HIV screening services that is peer led and located in the very center of the gay area of Barcelona, BCN Checkpoint is uniquely situated to provide a reference point to address the epidemic in the city. The main hypothesis driving this service was that increased HIV testing coverage, shorter time between HIV tests, and the introduction of tests to detect earlier infection (with immediate hospital referral to commence antiretroviral treatment) may lead to a decrease in HIV incidence. The objective of the current study was to describe the trends of HIV incidence during the period 2009–2017 and their relationship to sequential interventions in MSM and TW who attended this community center.
Methods
BCN Checkpoint was established in 2006 to provide a community-based peer-led venue for free HIV testing that both destigmatized and normalized testing with the aim of increasing testing coverage. As the epidemic evolved, the center has run three sequential three-year community engagement interventions starting in 2009. Innovative testing strategies were introduced prior to formal recommendations and included the use of point-of-care (PoC) tests, testing of sexual partners, increasing the frequency of testing, and additional testing for asymptomatic sexually transmitted infections (STIs).
Community interventions
The strategy of community engagement at BCN Checkpoint was characterized by three distinct campaigns between 2009 and 2017 that are briefly described below (Figure 1).
Poster campaigns.
Setting
At the first visit, clients were given a unique numerical identifier which was used for all clinical samples and information collected and was maintained for all subsequent visits. Cellular telephone numbers and e-mail addresses were also collected to avoid duplications and to send out reminders for routine clinic visits. Written informed consent was obtained to allow anonymized use of all data collected. Clients completed a baseline survey that included demographics, sexual health and history, and HIV risk information. All information collected was updated at subsequent visits. The HIV tests offered and testing platforms used evolved over time as detailed in Table 1. Services included counselling both for new HIV infection and other issues related to sex and STIs, rapid referral to HIV treatment clinics for newly-diagnosed individuals, and ongoing peer-led support groups for persons with positive HIV tests.
Methods of HIV detection, confirmation, and phase of infection across periods.
N/A: not available; PCR: polymerase chain reaction; WB: Western blot.
aFrom 1 January 2009 to 31 December 2009: Alere Determine™ HIV-1/2. Alere Ltd., Stockport, UK.
bFrom 1 January 2010 to 24 March 2016: Alere Determine™ HIV-1/2 Ag/Ab Combo. Alere Ltd., Stockport, UK.
cFrom 25 March 2016: Alere™ HIV Combo. Alere Ltd., Stockport, UK.
dFrom 4 March 2015. Xpert® HIV-1 Qual. Cepheid. Sunnyvale, Ca., USA.
eFrom 26 November 2015. Xpert® HIV-1 Viral Load. Cepheid. Sunnyvale, Ca., USA.
Regulatory and clinical data collection
Written informed consent was obtained for all information collected and testing performed. All data entered into the BCN Checkpoint Cohort database was anonymized with the use of a unique clinic number only. Between 2009 and 2011, data on demographics (gender, birth country, educational level), partner number, HIV testing history, drug use, and STI history were collected. Additional information on hepatitis A and B vaccination was added in 2012–2014, and in 2015–2017 questions specifically asking about Chemsex drugs, oral pre-exposure prophylaxis (PrEP) use, and symptoms suggestive of HIV seroconversion were added.
Statistical methods
A descriptive analysis was conducted for the following variables: number of subjects, visits, time in months between visits, sociodemographic characteristics (gender, age, partner’s HIV status), sexual behavioral patterns (number of partners, condom use, drug use while having sex), and PrEP use. Depending on the type or distribution of each variable, absolute frequencies (N) and percentages (%), interquartile range (Q1–Q3), or standard deviation(s) were calculated. This descriptive analysis was presented calculated for HIV-positive and HIV-negative individuals separately to allow investigation of the association between these two groups. Categorical variables were evaluated by Chi square test or Fisher’s test, while Student’s t-test or Wilcoxon test was used for continuous variables. This descriptive analysis was carried out for each period of time (2009–2011, 2011–2014, and 2015–2017). To understand the evolution of HIV prevention tools available in our center, the methods of detection and confirmation used as well as the description of the phases according to Fiebig staging have been described in the three periods. 2
A negative binomial multivariate mixed model analysis, with repeated-measurements and subject as random effects, was conducted to see the relationship between time between visits (in days) for each individual (dependent variable) and the period of time as exposure variable, adjusted for partner status and number of partners. Results are presented in terms of incidence rate ratios (IRRs) and 95% confidence intervals (CIs).
Finally, we calculated the global HIV incidence rate (2009–2017) and for three periods (2009–2011, 2012–2014, and 2015--2017). Inclusion criteria required individuals to have an initial negative HIV test in each period, to have at least two tests during the period, and no loss to follow-up during the same period. Individuals younger than 18 years and/or with a positive HIV test at their first visit for each period were excluded from the incidence estimation. Lost to follow-up was defined in each period as follows: if the difference in days between the last HIV test performed and the last day of the period was greater than the 90th percentile of the days between visits of the period for all subjects, they were considered a loss to follow-up. Incidence rate of HIV were the new cases of HIV (confirmed HIV-positive tests) divided by the sum of years of each subject in the cohort, with results shown per 100 persons-years including 95% CI.
Results
Between 2009 and 2017 a total of 25,916 persons accessed services at BCN Checkpoint with a total of 49,630 visits (Table 2). The number of persons varied between 2009 and 2011 (9768), 2012–2014 (8750), and 2015–2017 (11,369). The mean number of visits per person per year increased from 1.69 to 1.87 to 2.07, respectively, over the three periods (Supplemental Table 1). There was a concurrent reduction in the time period between visits from the first (9.28 months) to the second (8.26 months) and third (6.49 months) periods. A total of 1365 persons were diagnosed with HIV infection, 809 (59.3%) on their first clinic visit, and were linked to care within a median of 8 (p25–p75: 5–14) days. The highest percent of persons with a new positive initial HIV test was in the 2009–2011 period (71.35%), with a reduction to 52.25 and 56.92% in subsequent periods. HIV-uninfected clients were more likely to have more frequent visits than those with HIV infection across all time periods (p < 0.001) (Table 2 and Supplemental Table 1).
Summary of participant demographics and characteristics.
aQuestion not always included in the categories.
bQuestionnaire introduced in 2015.
cQuestions introduced in 2017.
The population was predominantly cisgender male (99.4%), Spanish nationals (62.4%) with at least a secondary educational level (96.2%) and a mean age of 34.8 years (95% CI: 34.7–34.9). Distribution of individuals according to educational level and nationality is presented in Supplemental Table 2. Almost equal numbers of individuals were single (47.5%) or with a steady partner (48.1%), with more clients partnered in the 2012–2014 period (52.15%), compared to 2009–2011 (47.47%) and 2015–2017 (45.39%). Most individuals had between two and nine sex partners during the last six months or since last visit (43.3%) with 14.2% having between 10 and 19 partners. The median time between clinic visits was 7.51 (Q1–Q3, 4.07–12.46) months and decreased with increasing numbers of sexual partners with individuals who had over 20 sex partners being seen most frequently every 6.03 months (Q1–Q3, 3.34–10.92) (Table 2 and Supplemental Table 1).
Just over one-fifth (21.6%) of individuals always used condoms with regular partners, and this increased to 53.0% with casual sex partners. The most common drugs used with sex were ‘poppers’ (12.86%) (amyl nitrite) followed by cocaine (11.20%) and ecstasy (5.10%). Stimulant drugs used in the context of sex (Chemsex) such as methamphetamine, gamma hydroxybutyrate (GHB), and mephedrone were used by 3.60, 4.40, and 1.40% of individuals, respectively. GHB use in all persons increased from 3.20 to 5.60% over the three time periods. Drug use with sex was the only consistent statistically significant difference over time between people living with HIV (PLHIV) and HIV-negative clients, with more use among PLHIV. Oral PrEP was only available in the last time period by enrollment in a study or by buying it informally from internet sources. In this context, 400 individuals verified use, with 93 (23.3%) attending regular STI and renal function monitoring. The BCN PrEP·Point clinic opened in November 2017 and consequently the program that follows informal PrEP users only started seeing these individuals during that year (Table 2).
The methods of HIV testing are shown in Table 1, with all results confirmed by Western blot. Viral load (VL) testing only became available in the 2015–2017 period. Fiebig staging of HIV infection showed that more individuals were diagnosed at earlier stages in later years. This difference is most marked between the 2012–2014 and 2015–2017 periods with an increase from 1.24% to almost 6.72% of individuals diagnosed at stage 1/2 corresponding to EIA-negative and RNA (± p24)-positive testing. During the same period, individuals diagnosed up to stage 5 – corresponding to less than 90 days of HIV infection – increased from 16.18 to 48.62%.
Table 3 shows a reduction in the time between visits of 9% between the second and the first period (IRR: 0.91, 95% CI: 0.89–0.93) and 17% (IRR: 0.83, 95% CI: 0.81–0.86) between the third and the first period, after adjusting for partner status and the number of partners. Furthermore, we observed that those with 20 or more partners had a 25% reduction (IRR: 0.75, 95% CI: 0.72–0.78) in time between visits compared to those without partners after adjusting for period and partner status.
Negative regression mixed model with days between visits as dependent variables across periods adjusted by partner status and number of partners.
IRR: incidence rate ratio.
The HIV incidence across all time periods was 2.11 per 100 person-years (95% CI: 1.94–2.31) (503 HIV confirmed, 23,980 person-years). The initial HIV incidence between 2009 and 2011 was 4.17 (95% CI: 3.53–4.93) (137 HIV confirmed, 3284 person-years) falling to 2.83 (95% CI: 2.44–3.30) (169 HIV confirmed, 5960 person-years) between 2012 and 2014, and finally to a rate of 1.57 (95% CI: 1.30–1.89) (115 HIV confirmed, 7315 person-years) between 2015 and 2017 which represents a 62% reduction (IRR: 0.38, 95% CI: 0.29–0.48) between the first and third study period (p < 0.001) (Table 4).
Overall HIV incidence rate and HIV incidence rate over three time periods between 2009 and 2017.
IQR: Interquartile range; IR: Incidence rate; IRR; Incidence rate ratio.*At least two follow-up visits
Discussion
There was a decrease in HIV incidence in the BCN Checkpoint Cohort despite increasing numbers of clinic attendees, clinic visits, and persons using stimulant drugs over the years analyzed. Interventions such as rapid testing, increasing testing frequency, partner testing, and close linkage to care/initiation of ART likely contributed to this decreased incidence in the cohort.
Several other factors were also likely implicated in HIV incidence during this time. These included the increased acceptance of earlier ART, simplification of regimens, improved tolerability, and increased drug potency and effectiveness.3–5 Oral PrEP was unlicensed in Spain during the period studied and so likely had a limited impact on incidence beyond those individuals enrolled in later years in demonstration projects, clinical trials, or those accessing medication by informal pathways.
Earlier in the HIV epidemic, testing was usually undertaken due to a specific risk behavior such as condom breakage or a condomless sexual encounter and so was frequently associated with anxiety during counselling and the wait for results in the established medical infrastructure. 6 The early introduction of PoC testing employed by BCN Checkpoint aimed to normalize the HIV test as part of regular sexual healthcare, give real time results, and in so doing de-stress the act of testing. This approach has subsequently been shown to be acceptable and comparable to formal testing and has resulted in more frequent testing among gay and bisexual men. 7 It is notable that the introduction of testing every 3–6 months at BCN Checkpoint from 2015 was two years ahead of similar recommendations from the United States Centers for Disease Control and Prevention. 8
The encouragement of partner HIV testing allowed regular partners to take the decision as to whether they would continue to use condoms as a couple or not – sanctioning and breaking previous paradigms of ‘condom always’ which often resulted in feelings of failure and guilt. For the partnership, testing together also facilitated the conversation around condom use, extra partnership relations, and a more open approach to managing HIV risk.
With the promotion of more regular testing and the use of HIV VL tests in the later years, BCN Checkpoint detected infection at increasingly early stages. These newly-positive individuals had the opportunity to rapidly commence ART and potentially preserve immune function while also allowing the individual to modulate sexual behavior during a potentially highly infectious stage.
Established gay organizations and public health agencies in Barcelona were initially concerned with the focus of HIV testing in gay men and peer performance of screening tests outside a formal medical infrastructure. These concerns have diminished over time as community acceptance, increasing attendance figures, and the clear community benefit have been established and verified by cohort data. Furthermore, subsequent studies have consistently supported community PoC testing to be equivalent to standard testing using parameters such as linkage to care (and time to undetectable HIV VL), 9 feasibility, 10 and acceptability and its ability to reach persons who had not previously tested. 11 HIV (and STI testing) at BCN Checkpoint was groundbreaking not only for Spain but also for Europe as a whole. Since then, it has become a model that has been implemented in a number of cities in Spain and across Europe.
There are clear limitations to the data presented. This is a retrospective review of a prospective cohort. The observations on HIV incidence may have been influenced by the evolving testing strategies employed by BCN Checkpoint but there is no way to confirm this impact. Furthermore, the population of BCN Checkpoint clients, although large, is not representative of all Barcelona MSM and TW who, for their own reasons, have chosen to seek HIV testing and other forms of sexual healthcare at other venues. To address this, strategies currently employed at BCN Checkpoint to increase coverage in the community include the use of outreach workers to engage with other community stakeholders and individuals to inform them about service provision and encourage service use. Finally, there may be an underestimation of the incidence rates due to loss to follow-up. The main reasons for this were related to behavior patterns, although we did not observe major differences between periods in those excluded according to risk factors’ prevalence, so making comparison between periods feasible.
The interventions at BCN Checkpoint to encourage HIV testing, increase testing frequency, encourage partner testing, and to detect early HIV infection and facilitate rapid referral for treatment were introduced rapidly as it became apparent that this would represent best practice. These strategies later appeared in standard HIV guidelines and it is reasonable to consider that these interventions provided an earlier benefit to the large numbers of MSM and TW in the city that choose BCN Checkpoint for HIV screening.
Supplemental Material
sj-pdf-1-std-10.1177_0956462420932424 - Supplemental material for Falling HIV incidence in a community clinic cohort of men who have sex with men and transgender women in Barcelona, Spain
Supplemental material, sj-pdf-1-std-10.1177_0956462420932424 for Falling HIV incidence in a community clinic cohort of men who have sex with men and transgender women in Barcelona, Spain by Michael Meulbroek, Albert Dalmau-Bueno, Jorge Saz, Giovanni Marazzi, Fèlix Pérez, Josep Coll, Héctor Taboada and Ferran Pujol in International Journal of STD & AIDS
Supplemental Material
sj-pdf-2-std-10.1177_0956462420932424 - Supplemental material for Falling HIV incidence in a community clinic cohort of men who have sex with men and transgender women in Barcelona, Spain
Supplemental material, sj-pdf-2-std-10.1177_0956462420932424 for Falling HIV incidence in a community clinic cohort of men who have sex with men and transgender women in Barcelona, Spain by Michael Meulbroek, Albert Dalmau-Bueno, Jorge Saz, Giovanni Marazzi, Fèlix Pérez, Josep Coll, Héctor Taboada and Ferran Pujol in International Journal of STD & AIDS
Supplemental Material
sj-pdf-3-std-10.1177_0956462420932424 - Supplemental material for Falling HIV incidence in a community clinic cohort of men who have sex with men and transgender women in Barcelona, Spain
Supplemental material, sj-pdf-3-std-10.1177_0956462420932424 for Falling HIV incidence in a community clinic cohort of men who have sex with men and transgender women in Barcelona, Spain by Michael Meulbroek, Albert Dalmau-Bueno, Jorge Saz, Giovanni Marazzi, Fèlix Pérez, Josep Coll, Héctor Taboada and Ferran Pujol in International Journal of STD & AIDS
Footnotes
Acknowledgments
We would like to acknowledge the clients, staff, and volunteers of BCN Checkpoint; the funders who support the work of this community-based organization; and to Dr Ross D Cranston for his advice on manuscript preparation.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
